Chapter 7 Processing Instruction and Structured Input.

Post on 26-Dec-2015

251 views 2 download

Tags:

transcript

Chapter 7

Processing Instruction and

Structured Input

In this chapter we explore:

The nature of input processing Processing instruction, a grammar

instruction that has structured input at its core

Research on processing instruction that demonstrates its effectiveness

A set of guidelines for developing structured input activities

Input processing Traditional instruction consisting of drills in which

learner output is manipulated and the instruction is divorced from meaning or communication is not an effective method for enhancing language acquisition.

What is needed is a new pedagogy of grammar instruction that takes as its point of departure what we know about how grammatical forms and structures are acquired.

This pedagogy needs to work with input and with the processes learners use to get data from that input.

Input processing

Of concern is input processing, how learners initially perceive and process linguistic data in the language they hear.

In input processing, learners might encounter their first problems in dealing with the properties of the new language.

We must come to some understanding of what input processing looks like.

Intake from input

Input processing is concerned with those psycholinguistic strategies and mechanisms by which learners derive intake from input.

Intake refers to the linguistic data in the input that learners attend to and hold in working memory during online comprehension.

Form

Research on input processing attempts to explain how learners get form from input while their primary attention is on meaning.

Form here is defined as surface features of language, although input processing is also relevant to syntax.

The most complete model

VanPatten (1996,2003b) presents the most complete model of input processing in SLA.

The role of working memory is important in this model since some of the principles are predicated on a limited capacity for processing.

Humans develop mechanisms that allow them to selectively attend to incoming stimuli. Without such mechanisms, there would be informational overload.

VanPatten’s Principles

Principle 1(P1). The Primacy of Meaning Principle. Learners process input for meaning before they process it for form. P1a. The Primacy of Content Words Principle. Learners

process content words in the input before anything else. P1b. The Lexical Preference Principle. Learners will

tend to rely on lexical items as opposed to grammatical form to get meaning when both encode the same semantic information.

P1c. The Preference for Nonredundancy Principle. Learners are more likely to process nonredundant meaningful grammatical form before they process redundant meaningful forms.

VanPatten’s Principles continued…

P1d. The Meaning-before-Nonmeaning Principle. Learners are more likely to process meaningful grammatical forms before nonmeaningful forms irrespective of redundancy.

P1e. The Availability of Resources Principle. For learners to process either redundant meaningful grammatical forms or nonmeaningful forms, the processing of overall sentential meaning must not drain available processing resources.

P1f. The Sentence Location Principle. Learners tend to process items in sentence initial position before those in final position and those in medial position.

VanPatten’s Principles continued…

Principle 2(P2). The First Noun Principle. Learners tend to process the first noun or pronoun they encounter in a sentence as the subject or agent. P2a. The Lexical Semantics Principle.

Learners may rely on lexical semantics, where possible, instead of word order to interpret sentences.

VanPatten’s Principles continued…

P2b. The Event Probabilities Principle. Learners may rely on event probabilities, where possible, instead of word order to interpret sentences.

P2c. The Contextual Constraint Principle. Learners may rely less on the First Noun Principle if preceding context constrains the possible interpretation of a clause or sentence.

Based on VanPatten (2003b)

Principle 1

That learners are driven to get meaning from input (P1) has a set of consequences, the first being that content words are searched out first.

A great deal of form that is meaning-oriented or has referential meaning may also be expressed by a lexical item or phrase elsewhere in the sentence.

This observation led to the creation of the construct communicative value (VanPatten, 1985b).

Communicative value

Communicative value refers to the meaning that a form contributes to overall sentence meaning and is based on two features. [+/- inherent semantic value] [+/- redundancy]

If meaning can be retrieved elsewhere and not just from the form itself, then the communicative value of the form is diminished.

Forms with [- semantic value], regardless of redundancy, contain no communicative value.

Nature of communicative

value The nature of communicative value

is important for input processing. The more communicative value a

form has, the more likely it is to get processed and made available in the intake data for acquisition (P1d).

Conversely…

The less communicative value a form has, the more likely learners are to “skip” it in the input.

For learners to process forms of little or no communicative value in the input, they must be able to comprehend an utterance such that the act of comprehension does not tie up all their attentional resources.

P1f

From P1f it is clear that learners perceive and process items in one position better than another.

For example, learners are much more likely to pick up question words and their syntax than object pronouns or the subjunctive, which tends to occur inside the sentence.

Word order

Input processing is also concerned with word order.

P2, the first noun principle, may have important effects on the acquisition of a language that does not follow strict subject-verb-object (SVO) word order.

A Juan no le gusta esta clase mucho. (literally, “To Juan is not very pleasing this class,” meaning “This class is not very pleasing to Juan.”

Erroneous input

Research has shown that learners do indeed encode such pronouns and noun phrases as subjects, thus delivering erroneous input to their developing linguistic systems.

They think that Juan is the subject. It is not that meaning is gotten elsewhere; it

is that meaning is not gotten at all or is gotten wrong.

The form-meaning connections are not only filtered, they are altered.

Summary

Research on input processing attempts to describe: What linguistic data learners attend to during

comprehension Which ones they do not attend to What grammatical roles learners assign to nouns How position in an utterance influences what gets

processed. Intake is grammatical information as it relates to

meaning that learners have comprehended (or think they have comprehended.)

A reminder…

As a reminder, input processing is but one set of processes related to acquisition

That learners derive some kind of intake from the input does not mean that the data contained in the intake automatically make their way into the developing mental representation of the L2 in the learner’s head (i.e., intake does not equal acquisition).

Rethinking grammar instruction: Structured

input We now have some idea of what

learners are doing with input when they are asked to comprehend it.

We can begin to develop a new kind of grammar instruction-one that will guide and focus learners’ attention when they process input.

Processing instruction

Processing instruction consists of three basic components: Learners are given information about a linguistic

structure or form. Learners are informed about a particular processing

strategy that may negatively affect their picking up of the form or structure during comprehension.

Learners are pushed to process the form or structure during activities with structured input- input that is manipulated in particular ways to push learners to become dependent on form and structured to get meaning.

Processing-oriented grammar instruction

InputIntakeDeveloping SystemOutput

Processing Mechanisms

Focused Practice

An example of relating processing strategies to

instruction: Verb morphology We turn to activities that focus learners’

attention on verb endings; the goal is for learners to use these morphological endings to comprehend tense rather than solely rely on lexical items.

After learners receive a brief explanation of how past-tense endings work, they might first practice attaching the concept of past time to verb forms in an activity such as the following.

Listening for time reference

Listen to each sentence. Indicate whether the action occurred last week or is part of a set of actions oriented toward the present.

1. John talked on the phone.2. Mary helped her mother.3. Robert studies for two hours.4. Sam watched TV.5. Lori visits her parents.

Structuring the input

Note that only the very ending encodes tense in the input sentence.

Lexical terms and discourse that would indicate a time frame are not present, thereby encouraging learners to attend to the grammatical markers for tense.

The input has been structured.

An example of relating processing strategies to

instruction: Adjective agreement

This time we focus on the following strategy: P1d. Learners are more likely to process meaningful grammatical forms before nonmeaningful forms, irrespective of redundancy.

Some features of language do not have inherent semantic or communicative value.

For example…

In the Romance languages, adjectives must agree in number and gender with the nouns they modify, but this feature of grammar contributes little or nothing to the meaning of the utterance in most cases.

In the following Spanish-language activity, learners’ attention is directed toward proper adjective form by a task in which the adjective endings must be attended to.

Who is it?

Listen to each sentence in which a person is described. Determine which person is being described and then indicate whether you agree or disagree.

David Letterman Madonna1. Es dinámica. (She’s dynamic.)2. Es comprensivo. (He’s understanding.)3. Es reservada. (She’s reserved.)

Remember that learners apply a first-noun strategy to determine subjects and objects of sentences (“who did what to whom”)

With the French causative, this leads to misinterpretation and nonacquisition.

In this activity, learners are pushed to process correctly; to be sure this happens, sentences with the noncausative faire (faire du ski, “to ski”) that involve two people are also included.

An example of relating processing strategies to instruction: The French

causative

Who is performing?

Listen to each sentence. Then answer the question.

1. Who cleans the room?2. Who packs the bags?[Teacher’s script] Read each sentence ONCE.

After each sentence, ask for an answer.1. Claude fait nettoyer la chambre à Richard.2. Marc fait les valises pour Jean.

Research on processing instruction

There has been ongoing research regarding the effectiveness of processing instruction.

An important part of this research has examined the relative effects of processing instruction versus those of traditional instruction.

The study that launched this research agenda is VanPatten and Cadierno (1993).

It is the most frequently cited study and has been the impetus for a number of replication studies.

Research questions

VanPatten and Cadierno sought to answer the following research questions: Does altering the way in which learners

process input have an effect on their developing systems?

If there is an effect, is it limited solely to processing more input or does instruction in input also have an effect on output?

If there is an effect, is it the same effect that traditional instruction has (assuming an effect for the latter)?

Focus of the research

VanPatten and Cadierno compared three groups of learners: A processing instruction group (number=27) A traditional instruction group (number=26) A control group (number=27)

The processing group received instruction based along the lines presented earlier.

The focus was word order and object pronouns in Spanish.

Who did what In the processing treatment, learners first received

activities with right or wrong answers (“Select the picture that best goes with what you hear”) followed by activities in which they offered opinions.

In the traditional group, learners received involving a typical explanation of object pronouns and the complete paradigms of the forms.

The control group did not receive instruction on the target structure and instead read an essay and discussed it in class.

Assessment

Assessment consisted of two tests: a sentence-level interpretation test and a sentence-level production test.

These were administered as a pretest, an immediate posttest, a two-week delayed posttest, and a four-week delayed posttest.

Assessment continued…

The interpretation test consisted of ten target items and ten distractors.

The production test consisted of five items with five distractors.

The interpretation group was based on an activity performed by the processing group (“Select the picture that best goes with what you hear.”)

The production test was based on an activity the traditional group performed (“Complete the sentence based on the pictures you see.”)

Results!

The pretests yielded no differences among the groups on the two tests prior to treatment.

In the posttesting phase, the processing group made significant gains on the interpretation test whereas the traditional and control groups did not.

On the production test, both the traditional and processing groups made significant gains but were not significantly different from each other.

The control group did not make significant gains.

Conclusions

Altering the way learners process input could alter their developing systems.

The effects of processing instruction are not limited to processing but also show up on production measures.

The effects of processing instruction are different from those of traditional instruction.

Two for one By being pushed to process form and meaning

simultaneously, learners with processing instruction not only could process better but also could access their newfound knowledge to produce a structure that never produced during the treatment stage.

Members of the traditional group learner to do a task, while the members of the processing group actually experienced a change in their underlying knowledge that allowed them to perform on different kinds of tasks.

Areas for future research

Are the effects of processing instruction (PI) generalizable to other structures?

Are the effects of PI due to different explicit information?

Are the effects of PI observable with different assessment tasks?

Are the effects of PI different from the effects of other types of instruction?

Do the effects of PI hold over time?

Are the effects of P1 generalizable to other

structures? Cadierno (1995) replicated the VanPatten and

Cadierno study using the Spanish preterite (past) tense as the target structure.

Again contrasting a control group, a traditional instruction (TI) group, and a processing instruction (PI) group, Cadierno measured the effects of treatment via two measures: An interpretation test (Is the sentence you’re

hearing present, past, or future?) A production test (writing sentences in the past)

Results

Cadierno’s results matched those of VanPatten and Cadierno exactly

On the interpretation test, the PI group improved significantly, but the other two groups did not.

On the production test, both the PI and TI groups improved significantly but were not different from each other.

Cheng’s study

In her dissertation, Cheng (1995) conducted a study with ser and estar, the two major copular verbs in Spanish.

She compared a control, a processing, and a traditional group in the use of copular verbs with adjectives as the target.

Her results mirrored those of the original VanPatten and Cadierno study.

Farley’s study

In another study, Farley (2001a) demonstrated the effects of PI on the Spanish subjunctive with noun clauses.

In his study he showed that participants who received PI made significant gains in both interpretation and production abilities with the subjunctive both in form and use.

Buck’s dissertation

Buck (2000) investigated the relative effects of PI and TI in the acquisition of the present continuous (versus the present progressive) in English by native speakers of Spanish.

“Bill is smoking a pipe” versus “Bill smokes a pipe.”

He results indicated greater gains for the processing group that were maintained over time on the interpretation test.

VanPatten and Wong

In one other study, VanPatten and Wong (2003) demonstrated that PI was superior to TI with the French causative.

They compared a control, a processing, and a traditional group and measured outcomes with an interpretation and a production test.

Their results were the same as the results of the original study.

Acquisition of verbal morphology

In another study involving the acquisition of verbal morphology, Benati (2001) compared PI, TI, and a control group using the Italian future tense as the target structure.

His results were similar to but not the same as those of the original study.

On the interpretation task, the PI group improved significantly, the TI group did as well, and the control group did not.

However, the gains made by the PI group were significantly greater than those of the TI group. PI>TI>C

Are the effects of PI due to different explicit

information? In VanPatten and Oikennon (1996), the

researchers compared three groups. One that received PI exactly as in the original

VanPatten and Cadierno study. Another that received the structured input

activities only, with no prior explicit information and no explanation during the activities

Another that received explicit information only, with no structured input activities.

The researchers used the same assessment tests as in the original study.

Results

Both the regular processing group and the structured input-only group improved significantly but were not different from each other.

The effects of PI are due not to the explicit information provided to learners but to the particular nature of the structured input activities.

Computer-assisted language learning (CALL)

Sanz and Morgan-Short (2003) set out to test whether explicit feedback is necessary or helpful to learners.

They tested four groups using the variables [+/- explanation] and [+/- explicit feedback].

All groups, regardless of the combinations of these variables, received the same structured input as practice.

The groups The first group was [+ explanation] (explicit

information about the language and how to process it in the input) and [+ explicit feedback] (telling learners not only whether an answer is correct or not but what the problem is if the answer is not correct0

The second group was [- explanation] and [- explicit feedback] (learners received structured input only, with indications only of whether their answers were right or wrong.

The groups continued…

The third group was [+ explanation] but [- explicit feedback].

The fourth group was [- explanation] but [+ explicit feedback].

Results The results showed that all groups improved

significantly on the three assessment tasks (interpretation and two production tasks- a sentence-completion task and a video-retelling task) from pre- to posttests.

The researchers found that no group was better than any other on any task.

Neither explicit information nor explicit feedback seemed to be crucial for a change in performance.

Practice in decoding structured input alone (as in the second group) seemed to be sufficient.

Benati’s study

In one other recent study, Benati (2003) reported similar findings with the Italian future tense.

He compared a regular PI groups with a structured input-only group and an explicit information-only group.

Benati’s results The explicit-only information group improved

slightly from pre- to posttest measures, but that both PI and structured input-only groups improved much more, and they improvement was not significantly different from each other.

Both treatments were significantly better than the explicit information-only group.

These results suggest a major if not causative role for the structured input activities of PI and only a minor role, if any, for explicit information.

Conclusions

Structured input appears to be the causative variable in processing instruction.

This means that explicit information is not important if the types of activities learners are engaged in actually push them to alter their processing strategies and make more or better form-meaning connections.

Two additional studies

Farley (2003)- In Farley’s study, the target item was the subjunctive in noun clauses in Spanish.

Wong’s study (2003) focused on negation of indefinite articles and partitives with avoir in French.

In both studies, the learners who received structured input only, without any prior explanation of the rule, made significant gains.

Are the effects of PI observable with different

assessment tasks? In VanPatten and Sanz (1995) the

researchers investigated the effects of PI as measured by three kinds of output tests.

They compared a PI group to a control group, using the same materials as in VanPatten and Cadierno.

Their output tests

Their output tests included not only a sentence-level test but also a question-answer test (based on pictures) and a video-narration test.

They administered the output tests in two modes: written and oral.

In the video narration, participants must provide all vocabulary, all syntax, and all grammatical features on their own, without any prompts.

Their results VanPatten and Sanz found that the control group did

not improve on any tests. The PI group improved significantly on the

interpretation test and on the sentence-level test in both modes.

Their gains were significant in the written mode but just missed significance in the oral mode.

In all tests, the PI participants performed better on the written tests than the oral.

It appears that the effects of PI are observable in a variety of output tests and are not limited to sentence-level tests.

Are the effects of PI different from the effects of other types of instruction?

Farley (2001a) compared the relative effects of PI with the effects of “meaning-based output” instruction (MOI).

Farley based the PI materials on P1b of VanPatten’s model.

Farley’s initial activities pushed learners to attend to subordinate clauses without main clauses [in Spanish] and had them indicate what the possible main clause could have been (or vice versa).

Farley’s activities had learners combine main and subordinate clauses to express doubt and belief about various people, places, and events.

MOI Unlike TI, MOI contains no mechanical drills

and is based on the tenets of structured output activities that were first mentioned in VanPatten and Cadierno (1993).

Participants might have heard (translated from Spanish) “I don’t think that dogs…” and on a sheet of paper would see “(to be) intelligent.”

They would then have to indicate what the person might be saying by using the correct verb form.

Farley’s procedure

Both the PI and the MOI groups had two days of instruction on the Spanish subjunctive.

Farley assessed outcomes using a pretest and posttest design, with one posttest administered one month after treatment.

The tests consisted of an interpretation test based on the PI materials and a production test based on the MOI materials.

Farley’s results

Farley’s results differed from the results of the previous studies comparing PI with TI.

His results showed that the PI and MOI groups improved significantly on both the interpretation and the production tests, with no difference between them.

Thus, PI was not superior to MOI, and neither was MOI superior to PI.

Another Farley study

In Farley (2001b), he used the same design, procedure, and target structure as in his 2001a study.

His results: Although both groups improved on the interpretation task, only the PI group maintained its performance on a delayed task.

The MOI group declined in performance. Thus, PI did prove to be superior to MOI in

the long run.

Do the effects of PI hold over time?

In the studies reported so far, the longest time delay in administering a posttest in any one study was four weeks.

As we saw with Lightbown’s 1983 study, after one year the effects of instruction wore off and learners were back where they had been at the beginning.

Long-term effects of PI

In VanPatten and Fernández (2003), the long-term effects of PI were studied.

An immediate posttest was given after instruction, and another one was given eight months later to the students who had continued on to the next semester and who had completed all phases of study.

Results When VanPatten and Fernández compared

the posttest results to the pretest results, they found that, as in all other studies, the students improved significantly on both measures.

Even though the scores dropped somewhat on the eight-month delayed test, the students were still significantly better at performing the tests than they were prior to treatment.

At least in this one study, the observed effects of PI seem to be durable.

Guidelines for developing structured input activities

Present one thing at a time. Keep meaning in focus. Move from sentences to connected

discourse. Use both oral and written input. Have the learner do something with the

input. Keep the learner’s processing strategies

in mind.

Present one thing at a time

Structured input must be delivered to the learner’s developing system in an efficient way.

Maximum efficiency is achieved when one function and one form are the focus at any given time.

Breaking up paradigms and lists of rules if useful for two reasons. It allows the explicit presentation and explanation of

the grammatical structure to be kept to a minimum. Breaking up a paradigm is more likely to result in

attention directed toward the targeted item.

Keep meaning in focus

Learners should not engage in the mechanical input activities of traditional grammar instruction.

Remember that input should be attended to for its message so that learners can see how grammar assists in the “delivery” of that message.

Activity:Looking for verb endings

Check off the statements you think are true based on what you know about your instructor.

He enjoys teaching. He enjoys watching the news at

night.He enjoys preparing exams. He enjoys correcting exams.

Analysis of activity

In the activity, the task cannot be performed without the learner making the form-meaning connections.

In order to indicate whether or not the sentence is applicable to the instructor, the learner must know what the sentence means and how the grammar encodes meaning in each.

Move from sentences to connected discourse

When we teach grammar via structured input activities, it is important to begin with sentences first, the shorter the better.

Shorter, isolated sentences give learners processing time, whereas in longer stretches of speech, grammatical form can get lost if the demands to process meaning overwhelm the learner.

Use both oral and written input

In activities, the learner should be provided with opportunities to hear and see the input.

Any combination of oral and written input within a single activity is fine.

Although all learners need oral input, some learners benefit from “seeing” the language and even claim they need to see it in order to learn it.

Keep the learner’s processing strategies in

mind Learners should focus attention during

processing on the relevant grammatical items and not on other elements of the sentence.

If one is teaching object pronouns in Spanish, it does little good to have each sentence begin with an explicit subject.

“Target” input sentences should begin with object pronouns, and subjects should be implied or positioned after the verb.

Types of activities for structured input

Structured input activities consist of two broad types: Referential activities are those for which

there is a right or wrong answer and for which the learner must rely on the targeted grammatical form to get meaning.

Affective structured input activities are those in which learners express an opinion, belief, or some other affective response and are engaged in processing information about the real world.

Structured input activities classified by

response type

Structured input

activities

Supplying information

Binary options

Ordering/ranking

Selecting alternatives

Matching

Surveys

Binary options

ACTIVITY: The Typical StudentBased on your experience, determine

whether the following statements are likely or unlikely.

1. A student who works part-time takes more morning classes.

2. An engineering student studies more than an art student.

Matching In matching activities, the learner indicates

the correspondence between an input sentence and something else: Matching a picture to an input sentence, etc.

For each sentence in column A, indicate to which activity in column B it is most logically connected.

Associations

COLUMN A COLUMN B

Alice… She…

1. Works part-time.

a. Goes to the gym.

2. Exercises five times a

week.

b. Studies every night.

Gets good grades.

c. Earns $5.00 an hour.

Supplying information

In information-supplying activities that provide structured input, learners don’t produce the grammatical item that is being taught but something else.

Unless he had to, (name)…. Would never eat _______. Would never watch ________. Would never go to _________. Would never spend money on ________.

Selecting alternatives

Select the phrase that best completes each statement about your instructor.1. As soon as he gets home, my instructor….a. reads the mail b. has a cocktail.c. plays with his dogd. something else

2. When it’s time for dinner, he…a. prepares the meal b. helps with the mealc. waits for the meald. orders a pizza

Surveys

In a survey, one or both of the following can happen: The learner responds to a survey item The learner elicits survey information

from someone else. Surveys can use a variety of the

response formats already discussed (binary options, supplying information, selecting from alternatives, and matching)

Signature searches

Signature searches are a type of survey activity in which learners mill around the room attempting to find people who can answer affirmatively to a particular statement.

Learners who ask questions get input because they must read the questions and know what they mean.

Signature searches place input in the hands of the learner.

Ordering and ranking

Another type of activity for structured input involves having learners order items either in terms of importance or likelihood or in terms of chronology.

Students are engaged in processing and reprocessing the verb forms, all the time connecting them to meaning as they work through the task of this kind of activity.

Summary of chapter

Viewed in as much detail as possible the nature of processing information

Examined the nature on structured input activities and how they push learners toward more optimal processing of language data

Reviewed research on processing instruction that demonstrates in effectiveness

Summary of chapter 7

Viewed in as much detail as possible the nature of processing instruction

Examined in particular the nature of structured input activities and how they push learners toward more optimal processing of language data

Reviewed research on processing instruction that demonstrates its effectiveness, an effectiveness that comes essentially from carefully constructed activities.